toxins 08 00193.pdf


Aperçu du fichier PDF toxins-08-00193.pdf - page 2/9

Page 1 2 3 4 5 6 7 8 9



Aperçu texte


Toxins 2016, 8, 193

2 of 9

has uniquely evolved for intraspecific competition in mammals [9,10]. It is notable that venoms are
more versatile compared to poisons, as they can be used for all the aforementioned functions (predation,
defence, and competition), whereas poisons can only be used for defence to deter predators [2].
Despite their evolution being perhaps driven by one particular function, venoms are typically
used multifunctionally; for instance, both for prey capture and defence [1,11]. In contrast, although
poison is used solely for defence in all tetrapod groups [2,7,8], there is variation in how such toxins
are acquired by the animal. The two alternative strategies here are biosynthesis, in which the animal
produces the toxins in a specialised “poison gland”, and sequestration, in which toxins are obtained
from an environmental source such as the diet [7,12]. These different means of acquiring toxins may
well have different consequences for their evolution, but comparisons between these strategies are
lacking in the literature.
Relatively little research effort has been devoted to understanding the broad-scale evolutionary
patterns of toxic weaponry in animals. Tetrapods provide an excellent case study to investigate
the evolution of toxic arsenal due to the diversity they present. Therefore, in this study we aim
to understand the macroevolution of venoms and poisons throughout the clade Tetrapoda using
a phylogenetic comparative approach, and will explore the phenotypic evolutionary patterns that
characterise these traits.
2. Methods
2.1. Data Collection
Phylogenetic trees were obtained from the literature for each clade of tetrapods included in this
study: amphibians [13], squamate reptiles (lizards and snakes) [14], mammals [15], and birds [16].
A single tree was available for each clade except birds, for which we used a random sample of 50 full
trees (each with 9993 species) obtained from www.birdtree.org.
We collated data on toxic weaponry from an extensive search of published literature and reliable
online databases (see Table S1) for as many species in the phylogenies as possible. We recorded the
presence/absence of toxins, presence/absence of particular types of toxic weaponry (poison and
venom), presence/absence of particular functions of the toxic weaponry (predation, defence, and
intraspecific competition), and the mechanism for acquisition of toxins (biosynthesis or sequestration).
Where we refer to “categories” in relation to analyses we mean the eight variables above: toxic
weaponry, the two types of weaponry, the three functions, and the two acquisition strategies.
In total, 21,535 tetrapod species were included in the phylogenies used (2870 amphibians,
4510 mammals, 4162 squamate reptiles, and 9993 birds). Of these, we were able to obtain data
on venoms and poisons for 19,161 (89%) species (858 (30%) amphibians, 4488 (>99%) mammals,
3833 (92%) squamate reptiles, and 9977 (>99%) birds). Data handling and analysis was conducted in
R v3.2.2 [17]. The complete dataset is available online (see Data Accessibility section).
2.2. Ancestral State Reconstruction
We estimated ancestral states for each of our categories using Bayesian stochastic mapping [18]
on each major clade separately. We ran these analyses in the R package phytools [19] based on an
“ARD” model, in which gains and losses can occur at different rates. The prior on the root state
was estimated using transition matrices within the function. We generated 1000 stochastic maps for
amphibians, mammals, and squamates which were used to reconstruct ancestral states. For birds, we
instead simulated 10 stochastic maps on each of the 50 phylogenies. This allowed us to incorporate the
uncertainty in this set of trees, but the size and number of the bird phylogenies exhausted available
computer memory and so we based our inference on the smaller number of 500 maps in contrast to the
1000 used for other groups. Nevertheless, this should still give adequate information for inference of
basic macroevolutionary patterns.